JMIR Publications

ABSTRACT

Background:

Working Alliance refers to an important relationship quality between health professionals and clients that robustly links to treatment success. Recent research shows that clients can develop an affective bond quality with chatbots. However, little research has investigated if chatbots with differently close social roles and the option of letting users choose the social role of a chatbot affect the perceived relationship with a chatbot.

Objective:

The goal of the present study was to answer the following research questions: Which interpersonal closeness design cues manifest the social role of a chatbot (RQ1)? How does a chatbot’s social role relate to the affective bond with the chatbot and users’ usage intentions (RQ2)? And, exploratively, how does an individual’s demographic profile (i.e., gender and age) (RQ3) and the option to freely choose the social role of a chatbot affect these evaluations (RQ4)?

Methods:

Informed by the Social Role Theory and the Social Response Theory, we developed a Design Codebook for Chatbots with Different Social Roles along an interpersonal closeness continuum. Based on this codebook, we manipulated a fictitious healthcare chatbot to impersonate one of four distinct social roles common in healthcare settings (INSTITUTION, EXPERT, PEER, and DIALOGICALSELF) and examined effects on perceived affective bond and usage intentions in a web-based lab experiment (N = 251, MAge = 41.15, 57% females). Participants were either randomly assigned (nno-choice = 202) or could freely choose to interact with one of these Chatbot Personas (nfree-choice = 49). Separate MANOVAs were performed (a) to analyze differences between the Chatbot Personas within the no-choice group and (b) between the no-choice and the free-choice group.

Results:

While the main effect of the Chatbot Persona on affective bond and usage intentions was insignificant (p = .868), we found differences based on participants demographic profiles: i.e., main effects for gender (p = .039, ηp2 = .115) and age (p < .001, ηp2 = .192) and a significant interaction effect of Persona and age (p = .011, ηp2 = .102). Participants younger than 40 years reported higher scores for affective bond and usage intentions for the interpersonally more distant EXPERT and INSTITUTION chatbots; Participants older than 40 years reported higher outcomes for the closer PEER and DIALOGICALSELF chatbots. The option to freely choose a Persona significantly benefited perceptions of the PEER chatbot further (e.g., affective bond: Mfree-choice = 5.28, SD = 0.89 vs. Mno choice = 4.54, SD = 1.10, p = .003, ηp2 = .117).

Conclusions:

Manipulating a chatbot’s social role is a possible avenue for healthcare chatbot designers to tailor clients’ chatbot experience to user-specific demographic factors and to improve their perceptions and behavioral intentions towards the chatbot. Our results also emphasize the benefits of letting clients freely choose between chatbots.